Significance: Samsung is to pay $310m for the main connectivity and location products of the UK’s CSR, and also gains useful patents. For its part, CSR will develop new products in emerging sectors such as enhanced audio, many focused on the car and home not the handset. The deal shows that, as smartphone components become increasingly integrated, chip suppliers like Samsung need to provide a complete platform, a trend which is spurring a wave of acquisitions to fill in gaps. That, in turn, makes it very hard for more specialized chip vendors to find a role in the mobile device market, and they need to identify new niches, often beyond the cellphone, where innovation still tops scale and integration.
Samsung’s $310m acquisition of many assets of UK chipmaker CSR epitomizes some key trends in the mobile device world. One, the increasingly tight integration of many components on and around the system-on-chip makes it tough for specialists to gain significant presence on handsets. CSR, once the dominant force in Bluetooth, has suffered as companies like Broadcom packed all kinds of radios onto their platforms. Two, that in turn means that vendors wanting to be in the front rank of the device silicon space must control and preferably own every element of the complex smartphone innards, and Samsung Electronics has several gaps to fill to create a full platform around its Exynos processor. Three, there are new functions emerging for mobile devices, which will require new and ‘ for now at least ‘ discrete chips. CSR now aims to chase some of those new opportunities, and will come up against a host of start-ups in areas like enhanced audio or positioning. Some of those components will, in time, become standard in the handset, and so the whole cycle will begin again.
Samsung is paying $310m for the core businesses of CSR (formerly Cambridge Silicon Radio), gaining valuable patents as well as the technologies ‘ IPR which could benefit its handset sister company too. The Korean giant will also invest $34.4m, for a 4.9% stake in the remainder of CSR. The smaller company will then focus its efforts on its newer and higher margin activities such as in-car systems, while offloading the basis of its success, its leading Bluetooth business to Samsung, along with Wi-Fi and GPS/location.
Although CSR led the Bluetooth market, it never repeated the same success when it expanded into Wi-Fi, GPS and other key elements of the mobile device. And as all those elements increasingly become tightly integrated with the baseband/processor SoCs, the firm faced the risk of being sidelined by the all-in-one platform providers like Qualcomm and Broadcom.
In its statement, it effectively acknowledged that providing core mobile chips is now a game for giants only. ‘Samsung’s platform presence in handsets, together with its investment resources, development expertise, technological breadth, and customer reach, will provide the Handset Operations unit and the associated employees with strong future prospects.’ It acknowledged the intensifying competition and squeezed margins in handset chips, saying: ‘Without a platform presence in handsets, it is likely that the pressure on gross margins and market share will persist.’
Samsung has its own challenges to become a mobile chip supplier in the league of Qualcomm. Apart from his powerful position in memory chips, its main offering is the Exynos processor. This is well-regarded and lives in the forefront of trends like quad-core, as well as retaining a sibling bond with Apple’s A4/A5 family (which Samsung initially designed and continues to manufacture). However, Exynos has two disadvantages ‘ handset OEMs’ suspicion of being reliant on the sister firm of their biggest competitor; and the lack of a full surrounding chipset like those offered by Texas Instruments or Marvell. Qualcomm bought Atheros partly to address this need, and Samsung can be expected to make a string of purchases to flesh out its own mobile platform.
Earlier this year it bought Swedish low energy Wi-Fi chip designer Nanoradio, which could give handsets battery life advantages. The company re-entered the M&A fray after a decade of non’participation in 1997 when it bought Israeli imaging chip specialist TransChip, and it has made several, usually small purchases since (as well as making a failed $5.85bn bid for SanDisk). Last year it added Liquavista, a Dutch maker of ereader display technology, and Grandis, a pioneer in STT-RAM memory. Former CEO Choi Gee-sung said last year that Samsung would step up the M&A pace, commenting: ‘As Samsung is involved in many business areas, it cannot do them all alone. It needs many partners’ ‘ though his focus seemed to be more on boosting software capabilities for the handset division.
However, a new thinking by the traditionally acquisition-shy giant may pave the way for a takeover on a grander scale in the chip business, to help propel Exynos into the league of SnapDragon. The troubled ST-Ericsson is inevitably mentioned from time to time, and would bolster Samsung’s baseband activities, as would Renesas Mobile, which has reportedly been up for sale since May. Samsung could also do with a strong Wi-Fi purchase ‘ while CSR’s Bluetooth and GPS capabilities are considered among the best in the world, it has been less strong in WLan. The big players are filling out their platforms ‘ Intel bought Infineon Wireless for its modems, Qualcomm snapped up Atheros, and Samsung needs to move nimbly to avoid being left behind.
CSR is a good start. Under the terms of the deal, Samsung gets ‘substantially all’ of the UK firm’s handset chip resources including 310 staff (including 25% of the R&D team). There is an IPR aspect to the deal, since Samsung gains 21 US/international patents (which will be licensed back to CSR) and a worldwide, perpetual, royalty-free, licence to other intellectual property used in the affected products. CSR has about 2,000 patents, and its CEO, Joep van Beurden, said in a conference call that many of these are ‘relevant to the area that Samsung wants to compete in’. He added: ‘We have great IP. That was one of the reasons Samsung was so interested in us.
CSR retains existing handset connectivity and location products and the associated revenue, and also the right to use future developments in markets ‘other than handsets and mobile devices’.
It will now focus mainly on these non-cellphone sectors, and on five technologies which it has identified as having high growth potential in mobile gadgets and beyond. These are audio (voice and music), automotive infotainment, indoor location, imaging, and ‘Bluetooth Smart’, an ultra-low power variant targeted at devices such as game controllers. The deal is due to complete in the fourth quarter of 2012.
These are the kind of areas where many smaller chip firms are now concentrating, hoping to set the gold standard for an emerging mobile function ‘ which increasingly may be focused on a car, home device or embedded gadget, not necessarily a phone. As the handset silicon falls into the hands of a few majors, even the large chip providers are looking to adapt their technologies for newer and more profitable niches. For instance, Nvidia and Texas Instruments are building special versions of their applications processors to run in-car consoles and dashboards. And improving sound quality on devices is also seen as an important way to lure OEMs ‘ CSR believes this, as does NXP, which is currently driving this field.
CSR has its own mixed track record in acquisitions. Most importantly, it got heavily into location awareness with the 2009 purchase of SiRF and last year it unveiled a platform based on that technology, which was to be the heart of its future growth. In fact, its SiRFstarV architecture underpinned two new offerings, one for mobile applications and one for in-car infotainment systems ‘ the latter will therefore remain an important part of the surviving business. The SiRFusion location platform promised a much improved user experience for a range of mobile applications, while the SiRFprimaII platform aimed to enhance the auto navigation experience.
Less successful was the takeover of image processing chipmaker Zoran last year, though the audiovisual activity it built around that purchase also remains within the rump firm. Indeed, the sale to Samsung is really an extension of a strategy outlined last November, to work around the cellphone sector rather than trying to carve out a position against it. At the time, it split itself into three units ‘ connectivity (now with Samsung), location (split between the two firms) and audiovisual. With this trio, CMO Kanwar Chadha said CSR could live without a position in the dominant electronics categories, phones and PCs, by controlling ‘platforms for almost everything else’, including in-car infotainment and navigation; digital imaging with the Zoran video SoC and image processing for cameras; wireless audio; and Bluetooth for home or industrial systems such as smart metering. It has outlined a full architecture for the car and will follow up with similar platforms in the other target areas. The strategy sounded optimistic eight months ago ‘ with Samsung’s cash and perhaps its custom for some of the newer products, CSR’s future now looks smaller, but brighter.
Samsung invests in Stoke:
Samsung’s Venture Investment arm has plowed $5m into Stoke, which specializes in data gateways and offload systems. This is part of its bid to become a significant RAN supplier at last in the 4G world. Samsung had never made much impact in wireless networks outside Korea, until it succeeded in dominating the Mobile WiMAX platform. That experience gave it inroads into operators such as Sprint, as well as strong field knowledge of TDD/OFDM networks, which it is turning to advantage particularly in the TDD flavor of LTE. Vikash Varma, CEO of Stoke, says partnerships with firms like his own are Samsung’s way to fill out its complete 4G platform. “Samsung has ambitions to be in the top three LTE suppliers,” he told LightReading, and that means working with many allies. Stoke’s packet data gateway will complicate the small cell strategy which underpins Samsung’s bid to be in the LTE mainstream. Stoke’s most famous strategic investors to date have been NTT Docomo and India’s Reliance Communications, and along with venture capital funds, it has raised a total of $97m since its first round in 2004. Its gateways perform a range of functions in 3G or 4G networks, often shifting processes which usually live in a central core closer to the edge of the system. These include securing IP traffic such as VoIP calls, aggregating traffic from many sites, acting as a femtocell gateway, and offloading traffic to the internet or Wi-Fi.